Overview
Gamma is an AI-native platform for creating presentations, documents, and websites, and it is hiring engineers into a competitive San Francisco talent market. The idea, which came from Gamma's marketing side, was deliberately unconventional: rather than post another job listing, put paid media behind a creator video — a 1:26 film starring creator and comedian Austin Nasso playing a Gamma engineer — and see whether paid spend could actually drive qualified engineering candidates to careers.gamma.app.
Winston Francois was brought in to run the experiment end to end. The mandate:
- Test whether paid media behind creator content could recruit engineers, not just generate awareness
- Drive qualified candidate traffic to Gamma's careers page across Meta and LinkedIn
- Move fast, read the signal honestly, and come back with a clear point of view on what to do next
The challenge
- An unproven premise: using performance paid media as a recruiting channel — behind a single creator video — was a first-of-its-kind experiment with no playbook to copy.
- A hard-to-measure outcome: the real result — a qualified engineering hire — lives in Gamma's Ashby applicant system, several steps downstream of any ad click.
- Attribution gaps: Meta's conversion tracking was unreliable during the window and recorded no clean conversion events, so ad-platform data alone could not tell the full recruiting story.
- Two channels, two jobs: Meta had to carry efficient reach and volume while LinkedIn served as professional proof and self-qualification — each judged on its own terms.
The approach
1. Built the campaign around one authentic creator asset
- Anchored everything to the Austin Nasso hero video — the creator playing a Gamma engineer, riffing on the day-to-day of the job — and cut it into micro-hooks, explainer clips, and CTA snippets.
- Ran a test phase (July 27 – Aug 2) to validate tracking and delivery, then took the Austin Nasso creative live Aug 3 and scaled through Aug 31.
- Drove all traffic to careers.gamma.app, with downstream attribution handled via Ashby UTM parameters rather than native ad-platform lead tracking.
2. Ran Meta for efficient, high-intent volume
- Structured a Landing-Page-View campaign (Eng_SF_Broad_LPV) and optimized against cost-per-landing-page-view with a <$0.50 flag threshold.
- Tested seven creative concepts cut from the hero film — from the full-length “Elevator” edit to 6–19s hooks like “Dropdown Menu” and “Are You Guys Hiring?”
- Protected budget behind the winners and flagged underperformers (the “Skiing & Pretzels” lifestyle cut) for pause against the test-phase kill threshold.
3. Used LinkedIn for professional proof and self-qualification
- Ran a parallel Website-Visits campaign (WF [Traffic] – Eng Recruiting) targeting SF software engineers.
- Weighted spend toward the cross-channel winners — “Are You Guys Hiring?” and “Big Tech vs Gamma” — and retired the higher-cost test baseline once tracking was validated.
- Read LinkedIn as a qualification channel, accepting a higher cost per click in exchange for a professionally self-selected audience.
4. Reported straight and recommended forward
- Delivered a combined channel report with clear benchmarks (30% hook rate, 1.5% CTR, <$0.50 cost-per-result) and called performance against them honestly.
- Named exactly what worked, what to scale, and what to kill — including the finding that long-form narrative outperformed short cutdowns for this audience.
- Flagged the Meta measurement gaps openly and recommended fixing tracking before the next phase rather than overstating attributed results.
The results
| Metric | Change | Impact |
|---|---|---|
| Meta cost per landing-page view | $0.41 | Live Austin Nasso ad set — inside the <$0.50 flag threshold |
| Hero video hook rate | 45.1% | “Elevator” full-length cut — 15 pts above the 30% benchmark |
| Meta CTR | 3.82% | More than 2.5× the 1.5% campaign benchmark (4.80% on the hero) |
| Landing-page views & clicks | 57,838 | 55,574 Meta LPVs + 2,264 LinkedIn clicks |
| Impressions | 2.08M | 804,781 reach on Meta at a 2.30 frequency |
| Qualified engineering candidate | Goal met | The experiment's target outcome, despite broken Meta attribution |
Meta was the clear performance driver: the live Austin Nasso ad set delivered landing-page views at $0.41 each, and the full-length “Elevator” hero film alone drove 41K+ landing-page views at a 45% hook rate and a 4.80% CTR — well above the 30% hook-rate and 1.5% CTR benchmarks set for the campaign. LinkedIn played its intended role as a professional-proof channel, contributing reach and clicks at a higher cost per action.
The outcome that mattered most sat downstream of the dashboards: the experiment set out to prove whether paid spend behind a creator video could surface qualified engineering candidates, and it did — producing a qualified applicant and meeting the goal Gamma set — even as the team flagged the Meta attribution gaps rather than papering over them. Most of what the campaign returned was a clear, honest read on how to pair creator content with paid media next time.
The winning creative
One 86-second film starring creator Austin Nasso — playing a Gamma engineer — anchored the entire campaign, and was the single best-performing asset across both channels. It was then cut into shorter concepts to test hooks and scale reach.
Top performer overall
“Elevator” — the full-length hero film
The long-form cut absorbed 71% of Meta's Austin Nasso spend and returned proportionally more. Long-form, character-driven creator storytelling outperformed every short cutdown — challenging the usual “short-form wins” assumption for this audience.
The cutdowns that scaled it
“This was one of our weirder experiments: put paid spend behind a creator video and see if it could actually drive engineering candidates. The Winston Francois team ran it end to end, moved fast, and were straight with us at every step about what was working and what they'd do differently. We hit the goal we set, learned a lot about pairing creator content with paid, and I'd happily work with them again.”Fiona TurkoCreator Partnerships Lead, Gamma
What we learned
- Narrative depth beat snackable hooks: across Meta, hook rate and CTR rose almost linearly with length up to ~90 seconds, and the full-length creator-led story outperformed every short cutdown.
- Fix measurement before you scale: the real recruiting outcome lives in the applicant system, so downstream attribution has to be wired cleanly before the next phase, not after.
- Creator content and paid media are complementary, not interchangeable: the same “hiring” hook that won on Meta also translated to LinkedIn, but each channel earns its keep on a different metric.
- Being straight about what didn't work is the deliverable: naming the attribution gaps and the concepts to kill is what makes an experiment worth running — and worth running again.
